AI Paper Atlas

Research map / 3D vision / Depth and stereo

Towards Interpretable Deep Networks for Monocular Depth Estimation

Zunzhi You, Yi-Hsuan Tsai, Wei-Chen Chiu, Guanbin Li

ICCV 2021 · 3D vision › Depth and stereo · approach: Dense depth prediction and completion · 15 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
assign a depth range to each unitResNet-50 MDE network · NYUD-V2depth selectivity (testing): 0.4286 -> 0.7529Improved
assign a depth range to each unitSENet-154 MDE network · NYUD-V2depth selectivity (testing): 0.4691 -> 0.7693Improved
assign a depth range to each unitResNet-50 MDE network · NYUD-V2RMS: 0.443 -> 0.422Improved
assign a depth range to each unitSENet-154 MDE network · NYUD-V2delta<1.25: 0.874 -> 0.882Improved
assign depth ranges instead of directly regularizing selectivity (ablation)ResNet-50 MDE network · NYUD-V2depth selectivity (testing): 0.6039 -> 0.7529Improved
assign depth ranges instead of directly regularizing selectivity (ablation)SENet-154 MDE network · NYUD-V2depth selectivity (testing): 0.5694 -> 0.7693Improved
assign depth ranges instead of directly regularizing selectivity (ablation)ResNet-50 MDE network · NYUD-V2delta<1.25: 0.857 -> 0.861No clear effect
correct unit responses using ground-truth depth averagesbaseline ResNet-50 · NYUD-V2delta<1.25: 0.849 -> 0.779Hurt
correct unit responses using ground-truth depth averagesinterpretable ResNet-50 · NYUD-V2delta<1.25: 0.861 -> 0.947Improved
apply method to CSPN depth completionCSPN · NYUD-V2 depth completiondepth selectivity (testing): 0.4213 -> 0.9475Improved
apply method to CSPN depth completionCSPN · NYUD-V2 depth completionRMS: 0.117 -> 0.119Hurt
apply method at layer upconv1 of DenseNet-161 modelDenseNet-161 MDE · KITTIdelta<1.25: 0.963 -> 0.959No clear effect

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy